Oct. IS, 1919 Relation of Nutrient Medium, to Plant A bsorption 1 1 1 



found might reasonably be related to the extent of the root system and 

 rapidity of diffusion. There is certainly no convincing evidence that 

 selective adsorption by the sand plays any important role, although the 

 exact relations still remain to be worked out. This can be accomplished 

 only by comparing a sufficient number of plants under identical aerial 

 conditions and with complete control of the solutions in all cases. Some 

 preliminary indications suggest that the paraffin seal used in sand 

 cultures may be slightly inhibitive to growth. This might be omitted 

 in experiments planned to solve such questions as the above. 



It will not be necessary to enlarge on the earlier discussion in this article 

 with regard to the effect of hydrogen-ion concentration. The critical con- 

 sideration of this phase of the work has already appeared elsewhere 

 (16-19). It will suffice to emphasize the inapplicability of titration 

 methods in determining the reactions of the nutrient solutions, and the 

 fact that permeability or absorption are influenced by the hydrogen-ion 

 concentration, although a Pg value of 5 has not been found to be inhib- 

 itive either in solution or sand culture nor in peat soils when other 

 inhibitive factors are absent. 



Finally, brief attention must be given to the methods of stating results, 

 since these may in themselves modify the interpretation or planning of 

 further experiments. The practice in these investigations has been to 

 state the composition of the solutions in terms of parts per million of 

 each ion. This seems to be the most logical scheme, since it is not possible 

 to determine in such complex solutions the exact nature and concentra- 

 tion of the undissociated molecules and ions present. In most cases the 

 dissociation values are high, and there is no satisfactory evidence to show 

 that the ionic concentrations are not chiefly concerned in absorption by 

 the plant. Certainly whatever the original salts used, the properties of 

 the solution are those of the ions and molecules formed after solution has 

 taken place. Thus we are of the opinion that it is desirable to compare 

 solutions on the basis of parts per million of the various elements or 

 radicles present rather than on that of molecular or osmotic fractions of 

 the salts used in preparing the nutrient solutions. The latter method 

 may lead to neglect of the factor of total supply of essential ions, as pointed 

 out before. 



In a recent article Tottingham (46) suggests as a result of some experi- 

 ments performed in his laboratory that even with highly dissociated 

 salts the particular salt used has an effect apart from the ions formed 

 in the solution. It seems, however, that the e\ddence presented in 

 support of this idea is insufficient. If a large number of plants had been 

 used and the significance of the results evaluated, as in the previously 

 mentioned work of Waynick (sj), it is not apparent thai the same con- 

 clusion would have been reached. 



In the present investigation concentrations have been expressed for 

 convenience in terms of atmospheres, derived from determinations of 



